Estudio de la deforestación de bosque tropical amazónico mediante un análisis multitemporal de imágenes RADAR (SAR).

The research’s aims is to analyze deforestation in the Amazonian tropical forest by using SAR images, taking as a starting point the significant problem related to forest cover loss due to anthropic activities such as the expansion of the agricultural, livestock or logging frontier. SAR images have advantageous characteristics when detecting changes in an area; therefore, a chain of pre-processing and data management was established to obtain deforested sites and calculate forest loss rates from 2016 to 2020. Two different methodologies were applied: the first developed by Vargas et al. (2019) with automatic deforestation detection using VH polarization on an annual basis; and the second through a change detection algorithm developed by Canty (2020) on Google Colab portal every two years. Each methodology subjected to validation yielded positive precision results, with Canty's algorithm (2020) being the methodology that best detects changes due to deforestation by managing two polarizations (VH-VV). Regarding the deforestation analysis, rates of around -0.85% were reported. However, the results obtained are slightly lower than those calculated with data from the environmental authority (-1.09% for the 2016-2018 period). Both methodologies indicate the great potential of SAR images in detecting deforestation in tropical forests.

• Autore principale: Latacunga Vega, Byron Humberto (author)
• Altri autori: Toaza Garcés, Indira Monserratte (author)
• Natura: bachelorThesis
• Lingua: spa
• Pubblicazione: 2021
• Soggetti: ALGORITMO; BOSQUE; CAMBIO; DEFORESTACIÓN; SISTEMAS DE INFORMACIÓN GEOGRÁFICA; FOTOGRAMETRÍA Y FOTOINTERPRETACIÓN
• Accesso online: http://dspace.unach.edu.ec/handle/51000/7550